After introducing Bard, an experimental, conversational, AI chat service, Google launched its latest (artificial intelligence) AI model called Gemini. Developed by Google DeepMind in collaboration with other Google teams, Gemini aims to revolutionize how people interact with technology.
A multimodal model, Gemini is designed to process different types of information, from text and code to audio, images, and videos. Demis Hassabis, CEO of Google DeepMind, described Gemini as “the most capable and general model we’ve ever built.”
“The transition we are witnessing with AI will be the most profound in our lifetimes, far surpassing previous technological shifts,” he said.
Generative AI comes to Google ads
Google’s AI-powered environmental insights platform adds the Philippines
“Gemini has the power to create unprecedented opportunities for people worldwide, fostering innovation and economic progress on an unprecedented scale,” said Sundar Pichai, CEO, Google.
Google ecosystem integration
Gemini 1.0 comes in three optimized sizes, Ultra, Pro, and Nano, each equipped with advanced reasoning capabilities. Its proficiency in extracting insights from vast data sets promises breakthroughs across diverse fields, from scientific discoveries to financial innovations.
Gemini will soon be integrated into Google products. Starting with Bard and Pixel, this technology will gradually become available across platforms like Search, Ads, Chrome, and Duet AI. Developers, too, will have access through Google AI Studio and Google Cloud Vertex AI, enabling them to harness its power for their projects.
According to Google, the Ultra version of Gemini surpasses human expert performance on the Massive Multitask Language Understanding (MMLU) benchmark, showcasing its exceptional reasoning prowess across various domains. Its native multimodality sets it apart, enabling seamless comprehension and analysis of different inputs.
Coding
One of Gemini’s standout strengths lies in coding tasks, excelling in understanding, explaining, and generating high-quality code across popular programming languages. Google said that its superiority extends to competitive programming, outperforming its predecessor, AlphaCode, by solving nearly twice as many problems.
“Google’s commitment to responsible AI development is evident in Gemini’s design,” the tech giant said.
Rigorous safety evaluations, including assessments for bias and toxicity, have been integrated into its development process. Stress testing, encompassing adversarial evaluations and risk assessments, ensures its safety across cyber-offense, persuasion, and autonomy domains.
Recognizing the importance of collaboration in ethical AI, Google is partnering with the tech industry to establish best practices and set the highest security and safety standards.
As Google DeepMind continues its innovative strides, the future holds promise with Gemini Ultra and subsequent versions, poised to offer even more advanced capabilities. The era of Gemini represents a transformative period for AI, with the potential to amplify creativity, expand knowledge, and reshape daily lives and work routines for billions worldwide.